Systems and methods for dynamic imaging review charters for clinical trials

Abstract

Systems and methods are provided for implementing dynamic imaging review charters (IRC) for clinical trials. One method includes accessing clinical trial data and annotation data; generating, using a artificial intelligence (Al) model, an electronic file related to the medical study, wherein the Al model generates the electronic file based on the published criteria data and the annotation data; and transmitting the electronic file to a client device configured to implement the electronic file in relation to performance of the medical study.

Description

SYSTEMS AND METHODS FOR DYNAMIC IMAGING REVIEW CHARTERS FOR CLINICAL TRIALSCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to U.S. Provisional Application No. 63 / 637,088, filed April 22, 2024, the entire contents of which is incorporated herein by reference.BACKGROUND

[0002] The performance of imaging-related clinical trials, such as, e.g., blinded independent central review (BICR) trials, involves conformance with a wide array of documents or standards, including, e.g., industry standards promulgated by the Food and Drug Administration (FDA) in the United Stated and regulatory agencies with similar responsibilities in other countries globally (collectively referred to hereinafter as “Regulatory7Agencies”), published oncology criteria (e.g., the response evaluation criteria in solid tumors version 1.1 (RECIST 1.1), the response assessment in neuro-oncology (RANG) criteria, the Lugano classification, the prostate cancer working group 3 (PCWG3) criteria, etc.), the study protocol experimental agent information, including mechanism of action, disease specific information, and other such criteria, standards, and information.

[0003] One document used during such trials is the imaging review7charter (IRC) prepared in accordance with the FDA’s Final Guidance for Industry7Clinical Trial Imaging Endpoint Process Standards. The IRC is a regulatory document that attempts to define and describe — in great detail — the procedures that an imaging core lab or imaging contract research organization (CRO) is to follow when processing imaging data or interpreting the resulting images (e.g., analysis of the images) to produce data toward a desired endpoint or similar data.

[0004] Present methods of generating such IRCs are challenging in a number of respects. For example, “standard”, published oncology7criteria and standards are often sub-optimal when actually applied to a particular trial, typically involving clarifications and / or modifications. Furthermore, it is common for the standards, criteria, trial goals, or endpoints to change during the course of the trial. As a result, it is arduous for humans to coordinate all available data, including evolving data from industry publications, to develop, maintain, and revise the IRC for optimal imaging evaluation in response to the large volume of data variable and changes that may be prudent as protocol and other changes occur.

[0005] Accordingly, there is a long-felt need for systems and methods that automate the design of IRCs for use in imaging-related clinical trials.

[0006] The discussion above is merely provided for general background information and is not intended to be used as an aid in determining the scope of the claimed subject matter.SUMMARY

[0007] Various aspects of the present disclosure relates, generally, to IRCs of the type used in connection with the performance of imaging-related clinical trials and, more particularly, to providing dynamic IRCs using machine learning or artificial intelligence and other computer- implemented, automated methods. In some examples, the technology described herein may relate to a unique, dynamic IRC system that advantageously provides intelligent, tailored modifications and clarifications to the standard (and suboptimal or unsatisfactory) published oncology criteria, thereby enhancing the quality of services provided while increasing the regulatory success record associated with image-related clinical trials.

[0008] In some configurations, the technology disclosed herein provides a system. The system may include one or more electronic processors. The one or more electronic processors may be configured to receive, over a communication network, a query to generate an electronic file for a medical study, where the query may indicate a plurality of characteristics related to the medical study. The one or more electronic processors may be configured to, responsive to the query, access published criteria data and annotation data corresponding to the published criteria data. The one or more electronic processors may be configured to generate, using a first artificial intelligence (Al) model, the electronic file related to the medical study, where the first Al model may generate the electronic file based on the published criteria data and the annotation data. The one or more electronic processors may be configured to transmit, over the communication network, the electronic file to a client device, the client device configured to implement the electronic file in relation to performance of the medical study.

[0009] In some configurations, the technology disclosed herein provides a method. The method may include receiving, with one or more electronic processors, a query to generate an electronic file for a medical study, where the query may indicate a plurality of characteristics related to the medical study. The method may include, responsive to the query, accessing, with the one or more electronic processors, published criteria data and annotation data corresponding to the published criteria data. The method may include generating, with the one or more electronic processors, using a first artificial intelligence (Al) model, the electronic file related to the medical study, where the first Al model may generate the electronic file based on the published criteria data and the annotation data. The method may include transmitting,with the one or more electronic processors, over a communication network, the electronic file to a client device, the client device may be configured to implement the electronic file in relation to performance of the medical study.

[0010] In some configurations, the technology disclosed herein provides a non-transitory computer-readable medium to store instructions that, when executed by a processor cause the processor to: receive a query to generate an electronic file for a medical study, where the query may indicate a plurality of characteristics related to the medical study; responsive to the query, access published criteria data and annotation data corresponding to the published criteria data; generate, using a first artificial intelligence (Al) model, the electronic file related to the medical study, where the first Al model may generate the electronic file based on the published criteria data and the annotation data; and transmit, over a communication network, the electronic file to a client device, the client device configured to implement the electronic file in relation to performance of the medical study.BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The following drawings are provided to help illustrate various features of examples of the disclosure and are not intended to limit the scope of the disclosure or exclude alternative implementations.

[0012] FIG. 1 schematically illustrates a system for implementing dynamic IRCs for clinical trials according to some examples.

[0013] FIG. 2 schematically illustrates a server of the system of FIG. 1 according to some examples.

[0014] FIG. 3 is a conceptual overview of an IRC document structure in accordance with some examples.

[0015] FIG. 4 illustrates an example of an IRC document in accordance with some examples.

[0016] FIG. 5 is a flowchart of an example method for implementing dynamic IRCs in accordance with some examples.DETAILED DESCRIPTION OF THE PRESENT DISCLOSURE

[0017] The disclosed technology is not limited in its application to the details of construction and the arrangement of components set forth in the following description orillustrated in the following drawings. Other examples of the disclosed technology7are possible and examples described and / or illustrated here are capable of being practiced or of being carried out in various ways.

[0018] A plurality of hardware and software-based devices, as well as a plurality of different structural components can be used to implement the disclosed technology. In addition, examples of the disclosed technology7can include hardware, software, and electronic components or modules that, for purposes of discussion, can be illustrated and described as if the majority of the components were implemented solely in hardware. However, in at least one example, the electronic based aspects of the disclosed technology can be implemented in software (for example, stored on non-transilory computer-readable medium) executable by at least one processor. Although certain drawings illustrate hardware and software located within particular devices, these depictions are for illustrative purposes only. In some examples, the illustrated components can be combined or divided into separate software, firmware, hardware, or combinations thereof. As one example, instead of being located within and performed by a single electronic processor, logic and processing can be distributed among multiple electronic processors. Regardless of how they are combined or divided, hardware and software components can be located on the same computing device or can be distributed among different computing devices connected by at least one network or other suitable communication link.

[0019] Currently, there is no widely recognized or established system for dynamically generating and updating imaging review charters (IRCs) using machine learning and automation in clinical trials (also referred to herein as medical studies). While IRCs are documents that outline procedures for processing and interpreting imaging data, the development and maintenance of IRCs are typically manual and time-consuming processes. The FDA provides guidance on clinical trial imaging endpoint process standards, emphasizing the importance of standardization and compliance with regulatory requirements. However, these guidelines do not specify the use of automated systems for IRC creation or updates.

[0020] The proposed technical solution for dynamic IRCs addresses a significant gap in current practices by leveraging machine learning and automation to tailor IRCs to specific clinical trials. The technology7disclosed herein advantageously enhances the efficiency and accuracy of IRC creation, ensuring that the IRC(s) remain aligned with evolving standards and trial specifications. While there are systems that support imaging processes in clinical trials, these systems primarily focus on managing and standardizing imaging data rather than dynamically generating regulatory7documents like IRCs. The development of a dynamic IRCsystem using machine learning represents an innovative solution that could improve the quality and regulatory compliance of clinical trials involving imaging endpoints. The automation of this process could streamline these efforts and improve quality and efficiency by reducing the reliance on manual updates.

[0021] The technical solution proposed by the technology described herein has demonstrated success in streamlining and improving the efficiency and quality of imaging data analysis, for various criteria, such as, e.g.. RECIST. RECIST is a widely accepted standard for assessing tumor response in oncology trials. The technology disclosed herein effectively supports such criteria by ensuring a tailored implementation. However, to fully realize the envisioned efficiency for any published criteria, continued fine-tuning of the models described herein is encouraged. This involves integrating more literature and datasets to enhance the adaptability and precision of the technology disclosed herein. By doing so, the technology disclosed herein can be optimized to handle a broader range of imaging endpoints, ensuring that technology disclosed herein remains aligned with evolving standards and trial specifications.

[0022] Access to additional literature and criteria would allow for the development of more sophisticated models that can adapt to different trial designs and imaging modalities. This would not only improve the accuracy of imaging data analysis but also enhance compliance with regulatory guidelines, such as those outlined by the FDA. Moreover, the ability to incorporate diverse criteria enables the technology disclosed herein to support a wider array of clinical trials, making the technology disclosed herein a valuable tool for sponsors and researchers seeking to leverage imaging endpoints in their studies.

[0023] FIG. 1 illustrates a system 100 for implementing dynamic IRCs according to some examples. As illustrated in the example of FIG. 1. the system 100 can include one or more client devices 105 (referred to herein collectively as “the client devices 105” and individually as “the client device 105”), one or more clinical trial databases 110 (referred to herein collectively as “the clinical trial databases 110” and individually as “the clinical trial database 110”), one or more medical imaging databases 115 (referred to herein collectively as “the medical imaging databases 115” and individually as “the medical imaging database 1 15”), one or more IRC databases 120 (referred to herein collectively as “the IRC databases 120” and individually as “the IRC database 120”), and one or more servers 125 (referred to herein collectively as “the servers 125” and individually as “the server 125”). In some examples, the system 100 can include fewer, additional, or different components in different configurationsthan illustrated in FIG. 1. For example, as illustrated, the system 100 includes one client device 105, one clinical trial database 110, one medical imaging database 115, one IRC database 120, and one server 125. However, in some examples, the system 100 can include fewer or additional client devices 105, clinical trial databases 110, medical imaging databases 115, IRC databases 120, servers 125, or a combination thereof. As another example, components of the system 100 can be combined into a single device or platform (e.g., the clinical trials database 110 and the medical imaging database 115 or the client device 105 and the IRC database 120). divided among multiple devices, or a combination thereof.

[0024] The client device 105, the clinical trial database 110, the medical imaging database 115, the IRC database 120, and the server 125 can communicate over wired or wireless communication networks 130. Portions of the communication networks 130 can be implemented using a wide area network, such as the Internet, a local area network, such as a Bluetooth™ network or Wi-Fi. and combinations or derivatives thereof. In some examples, the communication network 130 represents a direct wireless link between two components (or platforms) of the system 100 (e.g., via a Bluetooth™ or Wi-Fi link). Alternatively, or in addition, in some examples, two or more components (or devices) of the system 100 can communicate through an intermediary device of the communication network 130 not illustrated in FIG. 1.

[0025] In the example of FIG. 1. the clinical trial database 110 may store clinical trial data 135. The clinical trial data 135 can include information or data associated with one or more industry standards for conducting clinical trials (referred to herein as “industry' standard data’'), data associated with one or more templates for constructing an IRC (referred to herein as “IRC template data”), data associated with one or more published oncology criteria (referred to herein as “published oncology criteria data”), data associated with a particular clinical trial or study (referred to herein as “study data”), etc. As described in greater detail herein, in some instances, the clinical trial data 135 may be implemented (or otherwise used) to generate and maintain the dynamic IRCs as described herein.

[0026] In some configurations, the industry standard data may include a variety of documents, such as. e.g., Clinical Trial Imaging Endpoint Process Standards, Guidance for Industry (or simply “Final Guidance”), published by the U.S. Department of Health and Human Sendees and Food and Drug Administration (2018) or the International Council forHarmonization (ICH) of Technical Requirements for Pharmaceuticals for Human Use: ICH Harmonized Guideline Good Clinical Practice (GCP) E6(R3) (2023).

[0027] In some configurations, the published oncology criteria data may include, e.g., information or data related to Response Evaluation Criteria in Solid Tumors (RECIST), Response Assessment in Neuro-Oncology (RANG), the Lugano classification. Prostate Cancer Clinical Trial Working Group (PCWG), including applicable versions of each, and any other applicable standard now known or later developed. Such standards, while unsatisfactory' in a number of respects, provide a framework for developing customized IRCs as described herein.

[0028] In some configurations, the IRC template data may include a corpus of data from which entries may be selected to design a IRC applicable to a particular trial (including a variety of ‘'modifications” and ‘'clarifications” — collectively referred to as ‘'annotations” herein), as described in greater details herein. In some instances, the IRC template data may include a full range of options for annotations that are to be populated within an IRC. In some configurations, the IRC template data may be clarified or modified and augmented over time as the technology disclosed herein '‘learns” from past IRC designs as well as changes to the industry standards data, the imaging data 140, clinical trial endpoints, published oncology criteria data, etc. As such, in some configurations, both the IRC template data and the IRC may be dynamically adapted to such changes during the course of, and between the performance of, clinical trials, resulting in a valuable and novel IRC-related knowledge base.

[0029] In some instances, the clinical trial data 135 may include information or data specific to a particular clinical trial, such as, e.g., a particular clinical trial for which the IRC is being generated (e.g., the study data). As one example, the study data may include the study protocol and information regarding an experimental agent, such as, e.g., a mechanism of action of the experimental agent, data regarding prior studies related to the experimental agent, or other pertinent information as applicable. In some examples, the study data may include information or data related to a study protocol, a study drug, one or more disease indications, study contract or work order, etc. for a particular clinical trial. As one example, data associated with a disease indication may include an indication that the target disease is known to include both tumorous and non-tumorous lesions. As described in greater detail herein, in some instances, an indication that a target disease of the clinical trial includes non-tumorous lesions can be used by the technology disclosed herein to dynamically select or adjust annotations to published oncology criteria data used in a dynamic IRC.

[0030] As illustrated in FIG. 1, the medical imaging database 115 may store medical imaging data 140. The medical imaging data 140 may include one or more medical images or data. The medical imaging data 140 may include medical images captured using different modalities, such as, e.g., X-ray, computed tomography (CT), magnetic resonance imaging (MRI), ultrasound, position emission tomography (PET), single photon emission computed tomography (SPECT), etc. In some examples, after a medical image is captured using medical imaging equipment (e.g., an X-ray machine, a CT scanner, a MRI machine, etc.), the medical imaging equipment may transmit (e g., over the communication network 130) the medical image to the medical imaging database 115 for storage. In some instances, the medical imaging data 140 may be collected (or otherwise provided) as part of a clinical trial or study (or performance thereof). As described in greater detail herein, in some cases, the medical imaging data 140 may be analyzed or further processed using the dynamic IRC(s). Alternatively, or in addition, in some cases, the medical imaging data 140 may be generated (or otherwise captured) using the dynamic IRC(s) (such that the medical imaging data 140 is in compliance or aligned with the dynamic IRC(s)).

[0031] In some instances, as illustrated in FIG. 1, the system 100 may include the IRC database 120. In the example of FIG. 1, the IRC database 120 may store one or more IRCs 145 (e.g., one or more dynamic IRCs 145). As described in greater detail herein, in some instances, the IRCs 145 may be stored in IRC database 120 such that the IRCs 145 may be accessible to other components of the system. Alternatively, or in addition, in some configurations, the IRC(s) 145 may be stored by another component of the system 100, such as, e g., the client device(s) 105, the server(s) 125. etc.

[0032] In some configurations, the clinical trial database(s) 110, the medical imaging database(s) 1 15, the IRC database(s) 120, or a combination thereof may be implemented as a single database or multiple, distributed databases using a variety of database architectures and software. Alternatively, or in addition, in some configurations, the contents of the clinical trial database(s) 110. the medical imaging database(s) 115, the IRC database(s) 120, or a combination thereof (e.g., the clinical trial data 135, the imaging data 140, the IRC(s) 145, etc.) may be stored in a variety of formats and data structures, such as, e.g., unstructured text, structured text, associative arrays, spreadsheets, or another ty pe of data structure.

[0033] As illustrated in FIG. 1, the system 100 may include the client device(s) 105. The client device(s) 105 may be a computing device, such as, e.g., a desktop computer, alaptopcomputer, a tablet computer, an all-in-one computer, a notebook computer, a terminal, a smart telephone, a smart television, a smart speaker, a smart imaging device, a communication device, or another suitable computing device that interfaces with a user. The client device(s) 105 may be used by a user for interacting with or otherwise implementing a dynamic IRC (as described in greater detail herein). The client device 105 may also be referred to herein as a user device.

[0034] Although not illustrated in FIG. 1, the client device(s) 105 may include similar components as described herein with respect to the server 125, such as electronic processor (for example, a microprocessor, an application-specific integrated circuit (ASIC), or another suitable electronic device), a memory (for example, a non-transitory. computer-readable storage medium), a communication interface, such as a transceiver, for communicating over the communication network 130 and. optionally, an additional communication network or connection, and an HM1.

[0035] In some configurations, as illustrated in FIG. 1 , the client device(s) 105 may include at least one client application 150. In some cases, the client application 150 may be stored in a memory of the client device(s) 105. The client application 150 is a software application executable by an electronic processor of the client device(s) 105 in the example illustrated and as specifically discussed herein, although a similarly purposed module can be implemented in other ways in other examples. In some configurations, the client application 150 may be a dedicated software application locally stored in a memory of the client device(s) 105. The client application 150 (when executed by an electronic processor) may enable or facilitate implementation of or interaction with a dynamic IRC in accordance with the technology disclosed herein. As one example, a user may interact with the client application 150 in order to generate a dynamic IRC, as described in greater detail herein. As another example, a user may interact with the client application 150 in order to analyze or otherwise process the medical imaging data 140 (or a portion thereof) using the IRC(s) 145.

[0036] In the illustrated example of FIG. 1 , the client device 105 may include at least one display device 155 (referred to herein collectively as “the display devices 155” and individually as “the display device 155”). The display device 155 can provide (or output) a media signal or content to a user. As one example, the display device 155 can display a user interface (e.g., a graphical user interface (GUI)) associated with the client application 150. The display device 155 can be included in the same housing as the client device(s) 105 or can communicate with the client device(s) 105 over a wired or wireless connection. As one example, the displaydevice 155 can be a touchscreen included in a cellular phone, a smart wearable, a laptop computer, a tablet computer, a smart speaker, another type of portable smart device, etc. As another example, the display device 155 can be a monitor, a television, or a projector coupled to a terminal, a desktop computer, or the like via a cable.

[0037] The client device(s) 105 can include additional, different, or fewer components than those illustrated in FIG. 1 in various configurations. The client device(s) 105 can perform additional functionality other than the functionality described herein. Also, the functionality (or a portion thereof) described herein as being performed by the client device(s) 105 can be performed by another component, distributed among multiple computing devices (e.g., as part of a cloud service or cloud-computing environment), combined with another component, or a combination thereof.

[0038] As illustrated in FIG. 1, the system 100 may also include the server(s) 125. As illustrated in FIG. 2, the server 125 may include an electronic processor 200, a memon 205, and a communication interface 210. The electronic processor 200, the memory 205, and the communication interface 210 can communicate wirelessly, over a communication line or bus, or a combination thereof. The server 125 can include additional, different, or fewer components than those illustrated in FIG. 2 in various configurations. The server 125 can perform additional functionality7other than the functionality7described herein. Also, the functionality7(or a portion thereof) described herein as being performed by the server 125 can be performed by another component, distributed among multiple computing devices (e.g.. as part of a cloud service or cloud-computing environment), combined with another component, or a combination thereof.

[0039] The communication interface 210 can include a transceiver that communicates with one or more computing device(s) (e.g., the client device(s) 105), the clinical trial database(s) 110, the medical imaging database(s) 115. the IRC database(s) 120, another component of the system 100, another device external or remote to the system 100, or a combination thereof over the communication network 130 and, optionally, at least one other communication network or connection. The electronic processor 200 may include a microprocessor, an ASIC, or another suitable electronic device for processing data, and the memory7205 may include a non- transitory, computer-readable storage medium. The electronic processor 200 is configured to retrieve instructions and data from the memory 205 and execute the instructions.

[0040] As illustrated in FIG. 2, the memory 205 may include an IRC application 215 (referred to herein as “the application 215”). The application 215 may be a software application executable by the electronic processor 200 in the example illustrated and as specificallydiscussed below, although a similarly purposed module can be implemented in other ways in other examples. The electronic processor 200 can execute the application 215 to implement or otherwise facilitate the dynamic IRCs (e.g., the IRCs 145) described herein. In some examples, the electronic processor 200 may generate the IRC(s) 145 based on the clinical trial information 135, information provided via the client device(s) 105 (e.g., selections provided via a HMI of the client device(s) 105, as described in greater detail herein), etc. Alternatively, or in addition, in some configurations, the electronic processor 200 may facilitate a dynamic IRC generation process by dynamically guiding a client entity (or a user) through generation of a IRC (e.g., the IRC(s) 145). Alternatively, or in addition, in some configurations, the electronic processor 200 may monitor or otherwise track performance of a clinical trial in order to determine dynamic updates or adjustments to an IRC of the clinical trial, where the IRC was previously generated for the clinical trial (e.g., prior to starting the clinical trial).

[0041] In some configurations, as illustrated in FIG. 2, the application 215 may include (or otherwise implement) an IRC engine 220. The IRC engine 220 may be a software application or instructions executable by the electronic processor 200 in the example illustrated and as specifically discussed below, although a similarly purposed module can be implemented in other ways in other examples. In some instances, the electronic processor 200 may execute the IRC engine 220 as part of the application 215. For example, the IRC engine 220 may facilitate (or otherwise perform) dynamic IRC functionality as described herein, such as, e.g., generation (or creation) of a dynamic IRC, updating a dynamic IRC, etc. As described herein, in some configurations, the dynamic IRC may be custom (or specifically tailored) to a particular clinical trial.

[0042] In some examples, the IRC engine 220 may receive (or otherwise identity and retrieve), as input, the clinical trial data 135 (or a portion thereof), such as, e.g., the industry standard data, the published oncology criteria data, the study data, the IRC template data, etc. The IRC engine 220 may generate (or otherwise create) an IRC (e.g., the IRC 145). In some instances, the IRC engine 220 may generate the IRC based on the clinical trial data 135 (or a portion thereof). In some instances, the IRC engine 220 may generate the IRC using one or more Al or machine learning models, as described in greater detail herein. Accordingly, in some configurations, the IRC engine 220 may construct or update an IRC (e.g., the IRC(s) 145) based on the clinical trial data 135 (or a portion thereof), including, e.g., the industry standards data, the published oncology criteria data, the study data, the IRC template data, etc. The IRC engine 220 can retrieve data (e.g., the clinical trial data 135) from the clinical trial database(s)110 or other data source(s) (e.g., manually, via an automated query, etc.) and, in some instances, process the data using automated techniques, which may include employing a natural language processing model to extract relevant information for generating or updating the IRC (e.g., the IRC(s) 145). For example, as additional or updated clinical trial data are received (or otherwise become available), the technology disclosed herein can process the IRC(s) to dynamically adjust the IRC based on any changes to the clinical trial data 135.

[0043] As illustrated in FIG. 2, in some configurations, the memory' 205 may also include a learning engine 265 and a model database 270. In some configurations, the learning engine 265 develops one or more models using an artificial intelligence (Al) or machine learning function, algorithm, or model. Machine learning functions are generally functions that allow a computer application to learn without being explicitly programmed. In particular, the learning engine 265 is configured to develop a model based on training data. As one example, to perform supervised learning, the training data includes example inputs and corresponding desired (for example, actual) outputs, and the learning engine 265 progressively develops a model that maps inputs to the outputs included in the training data. As another example, to perform self-supervised learning (“SSL'’), a model is trained on a task using the data itself to generate supervisory signals (e.g., unlabeled training data), rather than relying on. e.g., external labels provided by a user (e.g., labeled training data). As yet another example, to perform semisupervised learning, the training data may include desired output values for a subset of the training data (e.g., labeled training data) while the remaining training data may be unlabeled or imprecisely labeled (e.g., unlabeled training data). Machine learning performed by the learning engine 265 may be performed using various types of methods and mechanisms including but not limited to decision tree learning, association rule learning, artificial neural networks, inductive logic programming, support vector machines, clustering, Bayesian networks, reinforcement learning, representation learning, similarity and metric learning, sparse dictionary learning, and genetic algorithms. These approaches allow the learning engine 265 to ingest, parse, and understand data and progressively refine models.

[0044] In some configurations, the learning engine 265 may develop one or more models configured to be implemented or otherwise facilitate the dynamic IRC functionality, as described herein. For instance, in some configurations, the learning engine 265 may develop a one or more models configured to, e.g., predict a desired annotation for a particular criterion, perform clustering (e.g., of similar IRCs), determine association rules, perform anomalydetection in a particular IRC, etc. Such models may perform classification (e.g., binary or multiclass classification), regression, clustering, dimensionality reduction, and / or such tasks.

[0045] Examples of such models may include, e.g., large language models (LLMs). such as GPT-x, Llama, etc. In some instances, the models may be fine-tuned via, e.g., reinforcement learning with RECIST 1.1 (or other criteria) compliance as a reward signal. Additionally, other transformer network systems known in the art, artificial neural networks (ANN) (such as a recurrent neural networks (RNN) configured to analyze temporal dependencies in protocol amendment histories and convolutional neural network (CNN)), decision tree models (such as classification and regression trees (CART)), ensemble learning models (such as boosting, bootstrapped aggregation, gradient boosting machines, and random forests applied for confidence scoring of generated inclusion criteria), Bayesian network models (e.g., naive Bayes), principal component analysis (PCA), support vector machines (SVM). clustering models (such as K-nearest neighbor. K-means. expectation maximization, hierarchical clustering, etc.), and linear discriminant analysis models. For IRC generation, these models may operate within a constrained architecture where LLM outputs are programmatically validated against protocol-defined logic gates (e g., tumor type, measurable disease at baseline, etc.), regulatory standards before finalization, etc.

[0046] In some examples, natural language processing techniques may be used when creating or updating the IRC(s) 145. As one example, data can be extracted from data sources (e.g., regulatory agency data sources, published criteria data sources, proprietary libraries of criteria clarifications and modifications, etc.) using a data-driven analytic methodology' to harness data and create the IRC(s) 145, a list of modifications to published criteria for use in the IRC(s) 145, a list of clarifications to published criteria for use in the IRC(s) 145, etc. As noted herein, in some instances, the clinical trial data 135 (or portion(s) thereof) may be stored as unstructured data (e.g., in a PDF format). In such instances, documents (e.g., the clinical trial data 135) retrieved from the clinical trial database 110 can be processed to convert the documents (e.g.. the clinical trial data 135) into a machine-readable text format. In some examples, such a conversion may be implemented using a multi-stage optical character recognition (OCR) pipeline combining Tesseract with custom image preprocessing, like bounding box detection for tabular criteria. As one example, the “pdftotexf ' package in Python can be used to read the documents (e.g., the clinical trial data 135) in a text format, and the "pvtesseracr package can be used to read the documents (e.g., the clinical trial data 135) in other formats. Various text cleaning and converting text to a bag of words can be implementedafter the documents (e.g., the clinical trial data 135) are read. In some instances, text embeddings (e.g., sentence bidirectional encoder representations from transformers (SBERT) fine-tuned on oncology trial protocols) can be generated for use with LLM(s) or other machine learning models (e.g., the model(s) developed by the learning engine 265). Thus, NLP and machine learning techniques can be used to automatically extract information or data from the documents (e.g., the clinical trial data) to be stored as industry standards data, published oncology criteria data. IRC template data, or a combination thereof. In some configurations, extracted data may undergoes semantic alignment checks against clinical data interchange standards consortium (CDISC) operational data model (ODM) templates to ensure interoperability with clinical trial management systems.

[0047] In some configurations, the technology disclosed herein may implement (or otherwise facilitate) a version control process or functionality that tracks IRC modifications across various iterations. In such configurations, the technology disclosed herein may implement a difference tracking model (e g., as a model developed by the learning engine 265), such as, e.g., the Myer’s Diff algorithm. For instance, the technology' disclosed herein may utilize the difference tracking model to generate an audit trail with timestamped change rationales synthesized by LLM(s) (e.g., "‘Protocol amendments updated per FDA guidance”).

[0048] In some configurations, the technology disclosed herein may implement discrepancy detection models (e.g., Siamese neural networks) to compare IRC versions against corresponding protocol amendments. In such configurations, the technology' disclosed herein may flag inconsistencies beyond a threshold (e.g., for manual review). For criteria involving expert validation, the technology' disclosed herein may generate and provide a human-in-the- loop interface to display LLM-generated options. In some instances, the LLM-generated options may be displayed alongside (or otherwise visually associated with) corresponding confidence scores. In some instances, the technology disclosed herein may implement gradient boosting functionality (e.g., extreme Gradient Boosting (XGBoost)) and one or more logistic regression models and supporting evidence snippets from published guidelines (e.g.. from the clinical trial data 135). In some configurations, approved modifications may trigger automated updates to IRC template libraries (e.g., the IRC template data of the clinical trial data 135) and backward-compatibility' checks against ongoing trials via graph-based dependency mapping.

[0049] As illustrated in FIG. 2, the models developed by the learning engine 265 may be stored in the model database 270 of the memory 205. Alternatively, or in addition, the modelsdeveloped by the learning engine 265 may be stored at another component or device external to the server 125. In such configurations, the models developed by the learning engine 265 may be accessible to the server 125 (or components thereof, such as, e.g., the IRC engine 220).

[0050] As noted herein, the server 125 (or component(s) thereof) can include additional, different, or fewer components than those illustrated in FIG. 2 in various configurations. Also, the functionality (or a portion thereof) described herein as being performed by the server 125 (or component(s) thereof) can be performed by another component (e.g., a remote computing device, another computing device, or a combination thereof), distributed among multiple computing devices (e.g., as part of a cloud service or cloud-computing environment), combined with another component (e.g., a remote computing device, another computing device, another component of the system 100, or a combination thereof), or a combination thereof. For example, in some instances, the application 215, the learning engine 265, the model database 270, or a combination thereof may be stored and executed by another component (e.g., either local to the server 125 or external to the server 125, such as, e.g., the client device(s) 105), distributed among multiple computing devices, combined with another component, or a combination thereof.

[0051] FIG. 3 schematically illustrates an example structure of the IRC 145 in accordance with some examples. In the example illustrated, the IRC 145 may include a portion that contains three columns. A first column may include an itemized list of parameters 305 relevant to the medical study (or clinical trial) (e.g., a first parameter 305A and a second parameter 305B, as illustrated in FIG. 3). A second column may include a list of published criteria 310 (e.g., the published oncology criteria data) (e.g., a first criteria 310A, a second criteria 310B, a third criteria 310C. a fourth criteria 31 OD. and a fifth criteria 31 OE. as illustrated in FIG. 3). A third column may include a set of annotations 315 (or annotation data) (e.g., a first annotation 315A, a second annotation 315B, athird annotation 315C, afourth annotation 315D, and a fifth annotation 315C, as illustrated in FIG. 3). In some configurations, the IRC 145 may be visually depicted (or otherwise represented) such that associations or relationships between components of the IRC 145 (e.g., a parameter, a criteria, and an annotation) are indicated (e.g., visually represented). For example, as illustrated in FIG. 3, the first parameter 305A is linked to the first criterion 310 A, the second criterion 31 OB, the third criterion 31 OC, and the fourth criterion 310D. Following this example, each criterion linked to the first parameter 305 A is further linked (or otherwise associated with) a corresponding annotation. For instance, the first criterion 310A is associated with the first annotation 315A, where the first annotation 315Amay be a modification or clarification (e.g., a revision) with respect to the first criterion 310A. Similarly, the second criterion 310B is associated with the second annotation 315B, where the second annotation 315B may be a modification or clarification (e.g., a revision) with respect to the second criterion 31 OB. As yet another example, as illustrated in FIG. 3, the second parameter 305B is associated with the fifth criterion 310E, which is further associated with the fifth annotation 315E (e.g., a modification or clarification (e.g.. a revision) with respect to the first criterion 310A).

[0052] As used herein, a modification may refer to changes to a specified (or corresponding) criterion. In some instances, a modification may not align with the specified criterion or an intent of that specified criterion. As used herein, a “clarification” may refer to one or more details of a criterion that is aligned with the criteria and with an intent of the criteria but may be listed as optional or not adequately specified in a corresponding publication (e.g., the published criteria data). As used herein, annotations (or annotation data) includes one or more modifications, clarifications, or a combination thereof.

[0053] FIG. 4 illustrates an example of an IRC document 400 (e.g., the IRC(s) 145) in accordance with some examples. The example illustrated in FIG. 4 is merely included to illustrate the nature of modifications and clarifications (e.g., annotations), and that a IRC (e.g., the IRC(s) 145) may include fewer or additional annotations (e.g., hundreds of such annotations).

[0054] In the example of FIG. 4, the IRC document 400 relates to a particular parameter, “target lesions” (represented in FIG. 4 by reference numeral 405). The IRC document 400 also includes criteria data 410 related to the target lesions parameter 405. In the example of FIG. 4, the criteria data 410 includes a set of criteria from RECIST 1.1 (e.g., the published criteria data). The set of criteria included in FIG. 4 includes criteria data 410 related to the selection of target regions based on size and involved organs, as well as the size of measurable non-nodal and nodal target lesions. The IRC document 400 also includes annotation data 415 (e g., a set of clarifications or modifications) of RECIST 1. 1 for the particular medical study related to the IRC document 400. For instance, the annotation data 415 includes a variety of modifications, clarifications, or a combination thereof to the published RECIST 1.1 criteria (e.g.. the criteria data 410 of FIG. 4). As one specific example, as illustrated in FIG. 4, the annotation data 415 includes an option to modify the medical study to increase the number of target lesions that can be selected, as well as the maximum of such lesions per organ. Following this examples, theannotation data 415 may include clarifications in accordance with three options for irradiated lesions.

[0055] FIG. 5 is a flowchart illustrating a method 500 for implementing dynamic IRCs according to some examples. The method 500 is described as being performed by the server 125 and, in particular, the electronic processor(s) 200 executing at least one of the application(s) 215, the IRC engine 220, the learning engine 265, the models of the model database 270, etc. However, as noted above, the functionality described with respect to the method 500 can be performed by other devices, such as one or more of the client device(s) 105, another remote server or computing device, another component of the system 100, or a combination thereof, or distributed among a plurality of devices, such as a plurality of servers included in a cloud sendee (e.g., a web-based service executing software or applications). Further, although generally described as being performed by a processor (e.g., the electronic processor 200), this processor may include multiple processors (e.g.. as apart of a distributed processor system or group of cooperating processors).

[0056] As illustrated in FIG. 5, the method 500 may include receiving, with the electronic processor 200, a query to generate an electronic file for a medical study (at block 505). In some configurations, the electronic file may represent an IRC or IRC document (e.g., the IRC(s) 145). Accordingly, in some instances, the IRC(s) 145 may be referred to herein as electronic files or electronic content.

[0057] In some configurations, the query may be generated at the client device 105. such as, e.g., by a user interacting with the client application(s) 150. In such configurations, the uery may be transmitted to the electronic processor 200 from the client device 105 over the communication network(s) 130. As such, in some instances, the query may be a user query' or request to generate the electronic file for a medical study.

[0058] In some examples, the query may indicate one or more parameters (or characteristics) related to the medical study for which the electronic file is to be generated for (e.g., study data). As one example, the query may indicate study data for a particular medical study, such as, e.g., information or data related to a study protocol, a study drug, one or more disease indications, study contract or work order, a mechanism of action of an experimental agent, data regarding prior studies related to an experimental agent, etc.

[0059] In some examples, the query may include the parameters (or characteristics) (e.g., the study data). Alternatively, or in addition, in some instances, the query may includeinformation or data that facilitates retrieval of the parameters (or characteristics) (e.g., the study data), such as, e.g.. from the clinical trial database 110.

[0060] The electronic processor 200 may retrieve the clinical trial data 135 (or a portion thereof) (at block 510). In some cases, the electronic processor 200 may retrieve the clinical trial data 135 responsive to receipt of the query. Alternatively, the electronic processor 200 may retrieve the published criteria data without receipt of the query'. In some instances, the electronic processor 200 may retrieve, as the clinical trial data 135, the published criteria data, the study data, the IRC template data, the industry standards data, or a combination thereof.

[0061] In some configurations, the electronic processor 200 may retrieve annotation data related to the clinical trial data 135. For example, in some configurations, the electronic processor 200 may retrieve (or otherwise access) the public criteria data and annotation data related to the public criteria data. In some examples, the annotation data may be preexisting data. The annotation data may be indexed or otherwise associated to various portions of the published criteria data (or other portions of the clinical trial data 135). As one example, various portions of the annotation data may be tagged as being associated with a corresponding position of clinical trial data 135. As one example, with reference to FIG. 3, the first annotation 315A may be tagged (or otherwise linked or indexed) to the first criteria 310A.

[0062] Alternatively, or in addition, in some configurations, the electronic processor 200 may generate the annotation data based on the clinical trial data 135 (e.g.. the public criteria data). For instance, the electronic processor 200 may provide the published criteria data (e.g., the clinical trial data 135), the query, or a combination thereof to an artificial intelligence (Al) model. In some instances, the Al model may be a model developed by the learning engine 265 and stored in the model database 270, as described in greater detail herein. In some examples, the Al model may be developed or trained to generate annotation data. In some configurations, the Al model may generate the annotation data based on the clinical trial data 135 (e.g.. the published criteria data, the study data, the IRC template data, the industry' standards data, etc.), the query', or a combination thereof. As such, in some examples, the Al model may generate annotation data related to the published criteria data, the query, or a combination thereof. As described herein, in some instances, the annotation data may indicate a revision (e.g., a modification or clarification) to the published criteria data (e.g., the set of criteria 310 of FIG. 3 or the criteria data 410 of FIG. 4). In some configurations, the Al model may generate (or otherwise determine) the annotation data, or the revision, based on at least one characteristic(or parameter) related to the medical study. For instance, the Al model may generate annotation data (or revision(s)) specific to the medical study (or a parameter thereof). Accordingly, in some instances, the electronic processor 200 may receive, from the Al model, the annotation data.

[0063] The electronic processor 200 may generate the IRC 145 (e.g., the electronic file) related to the medical study (at block 515). In some configurations, the electronic processor 200 may generate the IRC 145 (e.g., the electronic file or electronic content) based on the clinical trial data (e.g.. the published criteria data. etc.), the annotation data, or a combination thereof.

[0064] In some configurations, the electronic processor 200 may generate the IRC(s) 145 (e.g., the electronic file(s)) using Al-based techniques, user feedback-based techniques, or the like.

[0065] For instance, in some configurations, the electronic processor 200 may generate the IRC(s) 145 (e.g., the electronic file(s)) by implementing (or otherwise executing) one or more Al or machine learning models (e.g., as developed by the learning engine 265 or stored in the model database 270). For example, the electronic processor 200 may provide the annotation data, the clinical trial data 135. the query (or information included therewith), or a combination thereof to an Al model configured to generate an IRC (e.g., the IRC(s) 145). The Al model may be configured to generate the IRC(s) 145 based on the clinical trial data 135, the annotation data, the query (or information included therewith), etc. such that the IRC(s) 145 is custom (or tailored) to a particular medical study (e g., the IRC 145 is specific to a particular medical study). As such, in some instances, the electronic processor may receive, from the Al model, the IRC(s) 145 (e.g., the electronic file(s)).

[0066] Al-based techniques may include implementing Al or machine learning models, predictive analytics, or other techniques described herein. In some instances, the clinical trial data 135 may be input to one or more Al or machine learning models, which in turn may generate the IRC(s) 145 (or portions thereof). As one example, an Al or machine learning model may process the clinical trial data 135 and automatically generate trial-specific outputs for the IRC(s) 145. including, e.g.. one or more trial-specific modifications or clarifications to the published oncology criteria, trial-specific narratives, and the like.

[0067] Alternatively, or in addition, in some configurations, the electronic processor 200 may generate the IRC(s) 145 (e.g., the electronic file(s)) by implementing (or otherwise executing) a user feedback-based approach or technique. In such configurations, the electronicprocessor 200 may generate a graphical user interface (GUI) and transmit the GUI to the client device 105 over the communication network 130 for display to a user of the client device 105 (e.g., via the display device 155). In some configurations, the GUI may be displayed to a user of the client device 105 through the client application(s) 150, as described herein.

[0068] In some configurations, the GUI may visually present annotation options (e.g., potential annotations or revisions) (as visual or graphical representations). A user may interact wi th the GUI to select one or more annotation options (e.g., via one or more GUI elements or components). In some configurations, the annotation options may represent the annotation data (or portion(s) thereof). In some configurations, the electronic processor 200 may determine the annotation options (to be included in the GUI ) based on the annotation data, the clinical trial data 135. the query (or information included therein), etc. For example, the electronic processor 200 may determine the annotation option(s) (e.g., the set of annotations 315 of FIG. 3 or the annotation data 410 of FIG. 4) based on the parameter(s) of the medical study, the published criteria data, or a combination thereof. In some examples, the electronic processor 200 may invoke (or otherwise execute) one or more Al or machine learning models (e.g., as developed by the learning engine 265 or stored in the model database 270) in order to determine the annotation options, as described herein.

[0069] The electronic processor 200 may generate the GUI such that the parameter(s) of the medical study is visually associated with a corresponding criterion (or criteria) as well as a corresponding annotation option (or options) for that corresponding criterion (e.g.. as illustrated in and described with respect to FIGS. 3 and 4).

[0070] A user may interact with the GUI by selecting (or otherwise choosing) one or more of the annotation options included in the GUI. Accordingly, in some examples, the client device 105 may generate an electronic communication (or message) that includes selection data that may indicate the one or more annotation options selected by the user via the client device 105. The client device 105 may transmit the electronic communication (or message) over the communication network 130 to the electronic processor 200 (as, e g., a data packet). The electronic processor 200 may receive the electronic communication (or message) including the selection data from the client device 105. In some configurations, the electronic processor 200 may generate the IRC(s) 145 (e.g., the electronic file(s)) based on the selection data received from the client device 105. For example, when the selection data indicates a selection of a first annotation option, the electronic processor 200 may generate the IRC(s) 145 such that the IRC(s) 145 includes (or otherwise implements) the first annotation option.

[0071] Alternatively, or in addition, in some configurations, the electronic processor 200 may update the remaining annotation options based on the selection data. The electronic processor 200 may dynamically update the remaining annotation options in real-time (or near real-time). For instance, in some cases, a selection of a particular modification to the published criteria data may impact the available choices or options for other modifications to the published criteria data, trial configuration specifications for the imaging analysis, reader rules, etc. As such, the electronic processor 200 may dynamically update the annotation options available (e.g., included in the GUI) to reflect an impact of previously selected annotation options. Alternatively, or in addition, in some instances, the electronic processor 200 may update the IRC(s) 145 based on the selection data.

[0072] In some configurations, the electronic processor 200 may generate the IRC(s) 145 after a user-feedback approach or technique is completed. For example, a user may interact with a GUI component that indicates that the user is finished selecting annotation options (e.g., by clicking on a “submit” or “complete” button included in the GUI).

[0073] In some configurations, the electronic processor 200 may generate a complete IRC document, or one or more portions thereof. For example, the electronic processor 200 may generate or update a customizations and modifications table in the IRC(s) 145 (e.g., the set of annotations 315 of FIG. 3 or the annotation data 410 of FIG. 4). As described herein, such a table includes annotations (e.g., customization and modifications) to published oncology criteria used in the clinical trial (or medical study). Additionally, or alternatively, the electronic processor 200 may generate narratives for use in the IRC(s) 145 (or portion(s) thereof).

[0074] In some configurations, the electronic processor 200 may transmit the IRC(s) 145 (e.g., the electronic file(s)) to a remote device, such as, e.g., the client device 105, the IRC database 120, or the like (at block 520). The electronic processor 200 may transmit the IRC(s) 145 to the remote device such that the IRC(s) 145 may be implemented with respect to the medical study.

[0075] As one example, the electronic processor 200 may receive (or otherwise retrieve) the medical imaging data 140 (e.g., from the medical imaging database(s) 115). The electronic processor 200 may facilitate review or analysis of the medical imaging data 140 as part of the medical study based on (or in accordance with) the IRC(s) 145 (or portion(s) thereof).

[0076] As another example, the electronic processor 200 may transmit the IRC(s) 145 to the IRC database 120 for storage such that the IRC(s) 145 may be accessible, such as, e.g., bythe client device 105 or another device at a later point in time, such that the IRC(s) 145 may be implemented with respect to the medical study.

[0077] As another example, the electronic processor 200 may derive final imaging-based primary endpoints in accordance with the IRC(s) 145. In some configurations, another device or component (e.g., remote from the electronic processor 200) may derive final imaging-based primary endpoints in accordance with the IRC(s) 145), such as, e.g., the client device(s) 105. Alternatively, or in addition, the electronic processor 200 may derive (or otherwise generate) one or more additional (or secondary) electronic files (or documents) based on the IRC(s) 145. For example, the electronic processor 200 may derive atrial configuration specification (as an example secondary electronic file or document), which may be utilized (or otherwise implemented) for the design of image analysis software to be used during performance of the medical study. As another example, the electronic processor 200 may derive (or otherwise generate) an additional electronic file (or document) that includes reader rules used to train physicians that perform image analysis for the medical study.

[0078] The IRC(s) 145 (or portion(s) thereof) generated by the electronic processor (based on. e.g., the clinical trial data 135) can then be displayed to a user (e.g., via the display device 155 of the client device 105), stored for later use or further processing (e.g., in the IRC database(s) 120), or both, as described herein. As described herein, in some instances, the IRC(s) 145 can be used to further generate additional documents for the clinical trial (or medical study), including, e.g., trial configuration specifications, reader rules, and the like. As one example, the electronic processor 200 may transmit the trial configuration specification(s) to a computer system implementing radiology analysis software (e.g., such as the client device 105 or another device of the system 100). In such instances, the trial configurations specifications can be received by the radiology analysis software and used to guide analysis of images during the clinical trial (or medical study).

[0079] In some implementations, the electronic processor 200 may additionally perform a quality assurance (QA) technique on the IRC(s) 145 (or portion(s) thereof). As part of the QA technique, the electronic processor 200 can analyze the IRC(s) 145 for errors, such as, e.g., inconsistencies between the IRC narrative. IRC tables (e.g., customizations and / or modifications to the published criteria data), the clinical trial study protocol, or other related documents or data. When errors are identified, the electronic processor 200 can flag those portions of the IRC(s) 145 for review, such as, e g., by a user via the client device 105. In some implementations, flagging the errors can include determining one or more suggested changesto mitigate (or remove) the error and transmitting the one or more suggested changes to the client device 105 for display to a user (e.g.. via the display device 155). In still other implementations, when errors are identified, the electronic processor 200 may automatically revise the IRC(s) 145 (or portion(s) thereof) to resolve (or otherwise mitigate) the errors.

[0080] Accordingly, the technology disclosed herein provides a unique, dynamic IRC system that provides intelligent, tailored modifications and clarifications to the standard (and unsatisfactory) published criteria to enhance the quality of services provided as well as an increase success record from a regulatory point of view.

[0081] In some examples, aspects of the technology, including computerized implementations of methods according to the technology, can be implemented as a system, method, apparatus, or article of manufacture using standard programming or engineering techniques to produce software, firmware, hardware, or any combination thereof to control a processor device (e.g., a serial or parallel processor chip, a single- or multi-core chip, a microprocessor, a field programmable gate array, any variety' of combinations of a control unit, arithmetic logic unit, and processor register, and so on), a computer (e.g., a processor device operatively coupled to a memory), or another electronically operated controller to implement aspects detailed herein. Accordingly, for example, examples of the technology can be implemented as a set of instructions, tangibly embodied on a non-transitory computer-readable media, such that a processor device can implement the instructions based upon reading the instructions from the computer-readable media. Some examples of the technology can include (or utilize) a control device such as an automation device, a computer including various computer hardware, software, firmware, and so on, consistent with the discussion below. As specific examples, a control device can include a processor, a microcontroller, a field- programmable gate array, a programmable logic controller, logic gates etc., and other typical components that are known in the art for implementation of appropriate functionality (e.g., memory, communication systems, power sources, user interfaces and other inputs, etc.).

[0082] Certain operations of methods according to the technology7, or of systems executing those methods, can be represented schematically in the FIGS, or otherwise discussed herein. Unless otherwise specified or limited, representation in the FIGS, of particular operations in particular spatial order can not necessarily require those operations to be executed in a particular sequence corresponding to the particular spatial order. Correspondingly, certain operations represented in the FIGS., or otherwise disclosed herein, can be executed in different orders than are expressly illustrated or described, as appropriate for particular examples of thetechnology. Further, in some examples, certain operations can be executed in parallel, including by dedicated parallel processing devices, or separate computing devices configured to interoperate as part of a large system.

[0083] As used herein in the context of computer implementation, unless otherwise specified or limited, the terms ‘'component,” “system,” “module,” “block,” and the like are intended to encompass part or all of computer-related systems that include hardware, software, a combination of hardware and software, or software in execution. For example, a component can be, but is not limited to being, a processor device, a process being executed (or executable) by a processor device, an object, an executable, a thread of execution, a computer program, or a computer. By way of illustration, both an application running on a computer and the computer can be a component. A component (or system, module, and so on) can reside within a process or thread of execution, can be localized on one computer, can be distributed between two or more computers or other processor devices, or can be included within another component (or system, module, and so on).

[0084] Also as used herein, unless otherwise limited or defined, “or” indicates a nonexclusive list of components or operations that can be present in any variety of combinations, rather than an exclusive list of components that can be present only as alternatives to each other. For example, a list of “A, B, or C” indicates options of: A; B; C; A and B; A and C; B and C; and A, B. and C. Correspondingly, the term “or” as used herein is intended to indicate exclusive alternatives only when preceded by terms of exclusivity, such as “either,” '‘one of,” “only one of,” or “exactly one of.” Further, a list preceded by “one or more” (and variations thereon) and including “or” to separate listed elements indicates options of one or more of any or all of the listed elements. For example, the phrases “one or more of A, B. or C” and “at least one of A, B, or C” indicate options of: one or more A; one or more B; one or more C; one or more A and one or more B; one or more B and one or more C; one or more A and one or more C; and one or more of each of A, B. and C. Similarly, a list preceded by “a plurality of’ (and variations thereon) and including “or” to separate listed elements indicates options of multiple instances of any or all of the listed elements. For example, the phrases '‘a plurality of A, B, or C” and “two or more of A, B, or C” indicate options of: A and B; B and C; A and C; and A, B, and C. In general, the term “or” as used herein only indicates exclusive alternatives (e.g., “one or the other but not both”) when preceded by terms of exclusivity, such as “either.” “one of,” “only one of.” or “exactly one of.”

[0085] Although the present technology has been described by referring to preferred examples, workers skilled in the art will recognize that changes can be made in form and detail without departing from the scope of the discussion.

Claims

CLAIMSWhat is claimed is:

1. A system, the system comprising: one or more electronic processors configured to: receive, over a communication network, a query to generate an electronic file for a medical study, wherein the query indicates a plurality of characteristics related to the medical study; responsive to the query', access published criteria data and annotation data corresponding to the published criteria data; generate, using a first artificial intelligence (Al) model, the electronic file related to the medical study, wherein the first Al model generates the electronic file based on the published criteria data and the annotation data; and transmit, over the communication network, the electronic file to a client device, the client device configured to implement the electronic file in relation to performance of the medical study.

2. The system of claim 1, wherein the one or more electronic processors are configured to: provide the published criteria data to a second Al model, the second Al model trained to generate annotation data related to the published criteria data, wherein the annotation data indicates a revision to the published criteria data based on at least one characteristic of the plurality of characteristics related to the medical study; and receive, from the second Al model, the annotation data.

3. The system of claim 1, wherein the one or more electronic processors are configured to: derive, using the electronic file, a set of additional electronic files related to the medical study.

4. The system of claim 3, wherein the set of additional electronic files includes trial configuration specifications for an image analysis software tool to be implemented during the performance of the medical study.

5. The system of claim 1, wherein the one or more electronic processors are configured to: receive, over the communication network, medical imaging data collected as part of the performance of the medical study; analyze the medical imaging data based on the electronic file; and transmit, over the communication network, an output of the analysis of the medical imaging data to the client device.

6. The system of claim 1, wherein the one or more electronic processors are configured to: receive, over the communication network, updated published criteria data; and dynamically update, using the Al model, the electronic file based on the updated published criteria data.

7. The system of claim 1, wherein the electronic file is an imaging review charter for the medical study.

8. The system of claim 1, wherein the one or more electronic processors are configured to generate the electronic file by : generating a graphical user interface (GUI) based on the electronic file, wherein the GUI includes the plurality of characteristics, and, for each characteristic included in the plurality of characteristics, at least one corresponding criterion based on the published criteria data and at least one corresponding annotation based on the annotation data; receiving, from the client device over the communication network, selection data related to a first annotation of a first characteristic of the plurality of characteristics included in the GUI; and dynamically updating the GUI to reflect the selection data related to the first annotation of the first characteristic.

9. The system of claim 8, wherein the one or more electronic processors are configured to dynamically update the GUI to reflect the selection data related to the first annotation of the first characteristic by changing a corresponding annotation of a second characteristic of the plurality of characteristics to a different corresponding annotation.

10. The system of claim 8, wherein the one or more electronic processors are configured to execute a third Al model to dynamically update the GUI to reflect the selection data related to the first annotation of the first characteristic.

11. A method, the method comprising: receiving, with one or more electronic processors, a query to generate an electronic file for a medical study, wherein the query indicates a plurality of characteristics related to the medical study; responsive to the query, accessing, with the one or more electronic processors, published criteria data and annotation data corresponding to the published criteria data; generating, with the one or more electronic processors, using a first artificial intelligence (Al) model, the electronic file related to the medical study, wherein the first Al model generates the electronic file based on the published criteria data and the annotation data; and transmitting, with the one or more electronic processors, over a communication network, the electronic file to a client device, the client device configured to implement the electronic file in relation to performance of the medical study.

12. The method of claim 11. further comprising: providing, with the one or more electronic processors, the published criteria data to a second Al model, the second Al model trained to generate annotation data related to the published criteria data, wherein the annotation data indicates a revision to the published criteria data based on at least one characteristic of the plurality of characteristics related to the medical study; and receiving, with the one or more electronic processors, from the second Al model, the annotation data.

13. The method of claim 11. further comprising: deriving, with the one or more electronic processors, using the electronic file, a set of additional electronic files related to the medical study, wherein the set of additional electronic files includes trial configuration specifications for an image analysis software tool to be implemented during the performance of the medical study.

14. The method of claim 11, wherein the one or more electronic processors are configured to: receiving, with the one or more electronic processors, over the communication network, medical imaging data collected as part of the performance of the medical study; analyzing, with the one or more electronic processors, the medical imaging data based on the electronic file; and transmitting, with the one or more electronic processors, over the communication network, an output of the analysis of the medical imaging data to the client device.

15. The method of claim 11, further comprising: receiving, with the one or more electronic processors, over the communication network, updated published criteria data; and dynamically updating, with the one or more electronic processors, using the Al model, the electronic file based on the updated published criteria data.

16. A non-transitory computer-readable medium to store instructions that, when executed by a processor cause the processor to: receive a query to generate an electronic file for a medical study, wherein the query indicates a plurality of characteristics related to the medical study; responsive to the query, access published criteria data and annotation data corresponding to the published criteria data; generate, using a first artificial intelligence (Al) model, the electronic file related to the medical study, wherein the first Al model generates the electronic file based on the published criteria data and the annotation data; and transmit, over a communication network, the electronic file to a client device, the client device configured to implement the electronic file in relation to performance of the medical study.

17. The non-transitory computer-readable medium of claim 1 , wherein the instructions, when executed by the processor, cause the processor to: receive medical imaging data; anddetermine, using the electronic file, one or more imaging endpoints based on the medical imaging data.

18. The non-transitory computer-readable medium of claim 16, wherein the instructions, when executed by the processor, cause the processor to: derive, using the electronic file, a set of additional electronic files related to the medical study, wherein the set of additional electronic files includes trial configuration specifications for an image analysis software tool to be implemented during the performance of the medical study.

19. The non-transitory computer-readable medium of claim 16, wherein the instructions, when executed by the processor, cause the processor to: generate a graphical user interface (GUI) based on the electronic file, wherein the GUI includes the plurality of characteristics, and, for each characteristic included in the plurality of characteristics, at least one corresponding criterion based on the published criteria data and at least one corresponding annotation based on the annotation data; receive, from the client device over the communication network, selection data related to a first annotation of a first characteristic of the plurality of characteristics included in the GUI; and dynamically update the GUI to reflect the selection data related to the first annotation of the first characteristic.

20. The non-transitory computer-readable medium of claim 19, wherein the instructions, when executed by the processor, cause the processor to: dynamically update the GUI to reflect the selection data related to the first annotation of the first characteristic by changing a corresponding annotation of a second characteristic of the plurality of characteristics to a different corresponding annotation.

Images